1. Field of the Invention
The present invention relates generally to tape drive systems. More particularly, the present invention relates to a system and method for predicting capstan slip conditions using the steady-state tape velocity corresponding to normal forces and drive forces at risk of slip in tape systems utilizing belt-driven tape cartridges.
2. Description of Related Art
Magnetic tape recording is used in a multitude of applications, ranging from audio and video recording to digital information recording. Computer systems utilize magnetic tape recording for a variety of purposes, including high-density data backup, on-line storage, archiving, versioning, multimedia edit and playback, Internet downloading, and application storage. Magnetically recorded tape cartridges are often used for the convenience of portability. The high storage capacities of today's tape cartridges provide advantages over other portable storage media, such as floppy disks, particularly due to the greatly increased size of software packages, graphics files, and the like.
Various types of tape cartridges exist, depending largely on the environment in which it is to be used. Typical audio tape cartridges, or cassettes, are relatively simple in their design, but are not capable of operating at high speeds with very high information densities. On the other hand, larger, more sophisticated tape cartridges can be used for data-intensive purposes, such as high-speed computer backups. One such tape cartridge is the quarter-inch cartridge (QIC). QIC technology involves methods of storing and managing information using QIC drives and media. This type of tape cartridge utilizes a belt-driven tape feed mechanism, and provides more precise read/write head alignment and managed tape tension and control. Indicia on the tape can be used to provide information to the tape drive during operation.
While magnetic tape cartridges remain popular, they are not without drawbacks. Digital information such as a program file is particularly sensitive to data loss, as the loss of any bits may be critical. Tape cartridges incorporate a great deal of mechanical technology, but mechanical parts can become defective over time or due to manufacturing defects. For example, tape drives move the tape in belt-driven tape cartridges past the read/write head by way of turning the tape reels with an elastomeric drive belt within the tape cartridge. The drive belt moves as a cartridge wheel turns, which is pressed against a rotating drive roller having elastomeric properties. If the friction between the cartridge wheel and the drive roller is not great enough, the drive roller can slip and spin against the cartridge wheel without moving the tape. When this occurs, the elastomeric material on the drive roller wears in an uneven fashion, which can lead to a failure of the tape drive. The cartridge wheel can also wear unevenly in such an instance, which can destroy the cartridge and result in data loss.
Due to the potential for mechanical malfunctions of tape cartridges, it would be advantageous to detect capstan slippage to notify the user of the problem as soon as possible. Prior art systems have attempted to detect actual capstan slip by monitoring for an abrupt change in tape velocity. However, these systems are only capable of determining a change in velocity of the tape at a particular time, and do not take into account normal velocity variations that are characteristic of belt-driven tape cartridges that have elastomeric properties in the drive roller and drive belt. Further, monitoring for a velocity change over a very short distance of the tape is actually tantamount to monitoring for actual capstan slipping. The prior art "detects" capstan slip by measuring the change in the tape velocity, where a "change" of a certain degree would indicate that the tape/capstan relationship has entered into a sliding friction zone, i.e., that slipping is occurring. However, capstan "slip detection" of this type does not intervene at a time prior to the potentially catastrophic slip condition, as the damage may have already occurred when the capstan slip is detected.
It is therefore desirable to "predict" capstan slip. One prior art system, described in U.S. Pat. No. 3,320,600, issued on May 16, 1967 to Headrick et al., provides a circuit which attempts to detect a forthcoming failure in capstan operation on a tape drive. Headrick et al. describes a system for determining a tape speed ratio by recording two indications in sequence with a particular time spacing. It measures a time interval between two inserted indications on the tape during writing, and compares it to a time between the two inserted indications during reading to calculate a change in instantaneous velocity. However, Headrick et al. does not predict capstan slip as in the present invention, but rather focuses on the failure of capstan operation itself, i.e., "capstan acceleration or deceleration response". Therefore, Headrick et al. does not determine whether slipping may occur due to varying force factors between the tape drive and the tape itself. Further, Headrick et al. is directed to a change in velocity based on tape length variables, which may not correspond to capstan slip, particularly in belt-driven tape cartridge systems.
Therefore, a method for predicting a capstan slip situation for belt-driven tape cartridges using the steady-state velocity, rather than a sudden deviation in tape velocity, is desired. The elastomeric properties of belt-driven tape cartridges naturally cause some velocity variations which are independent of capstan slip. Prior art systems which monitor a change in velocity may trigger capstan slip errors in cases where the change in velocity is normal due to normal velocity variations of belt-driven tape cartridges. It would also be advantageous to provide continuous prediction of potential capstan slip situations, rather than only during periods of tape acceleration or deceleration. The present invention overcomes these and other shortcomings, and provides additional advantages over the prior art.